xAI Jailbreak Evaluation: Results as Published

xAI's Grok 4.7 model card (section 8.1, Jailbreaks) reports how often the model complies with attacks that should be refused. It has three parts: a battery of standard jailbreaks, both published and internally discovered; StrongReject, the StrongREJECT suite; and long-horizon multi-turn jailbreaks, including Crescendo attacks. Each part is reported as compliance, the share of should-refuse attacks answered with disallowed help, and lower is better.

Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.